Outsourcing Innovations: The Pros And Cons

01 December 2010 |

Some carriers are thinking the once unthinkable and outsourcing their research and development to specialist firms. Tim Phillips weighs up the choices they must make.

Innovation is expensive and risky. It doesn’t come with a deadline attached. It can go nowhere. It is highly dependent on a resource that is hard to secure, scarce and variable in quality – innovative people. Badly done, it is everything that investors loathe. Spending on research and development (R&D) can be seen as a cost, and a high cost too. On the other hand, innovation is the main source of competitive edge for carriers. It can sustain that advantage, create customer loyalty and drive a share price. So, done well, it is everything that investors love. It is also essential. So the question is, who is best placed to do your innovation?

As the business model for carriers more closely approximates that of the IT industry, so more are feeling comfortable in outsourcing their R&D, often to firms whose reputation was made by doing similar jobs in the IT business.

Sharing the investment

Infosys has a dedicated R&D unit called Setlabs. Hema Prem, head of Europe for Setlabs, has done many jobs for carriers, some of which involve core development projects. “Outsourcing R&D is no longer a closed process that is performed in isolation,” he says. “In a mature market, involving the expertise of partners, vendors, academics and research organisations in the innovation process is inevitable.”

Infosys calls this “Innovation Co-creation”, and it has involved projects with customers such as Wireline, BT and Cable & Wireless in the US and Europe. For core technology, the carrier retains the intellectual property (IP), but for non-core projects, the investment and the IP can be shared.

The company claims that it wins customers simply because it is better at innovating than the companies it works for. “Setlabs research found that the maturity level of our existing IP was much higher than the subsequent six months of closed door R&D efforts – this is compared with a telco we were working with for the same area. After discussion with the telco it made sense to leverage our existing IP and decrease the time to market,” explains Mukul Gupta, head of communications, media and entertainment EMEA, at Infosys.

Many projects involve important but non-core development. For example, Setlabs worked with BT to develop a field optimisation suite. Pooling IP, the companies created an automated resource management application to forecast, plan and position a large field workforce.

Infosys claims a higher level of maturity in its technology, offering better insight and value than insourced R&D. It believes, in Gupta’s words, that: “There is merit in joining hands with partners for an open innovation in R&D and extending its R&D arms.” 

Not for everyone 

Not all carriers want to extend their arms, though – at least not for the type of high-value transformative innovation that’s taking place in the move to next-generation networks. Matthew Finnie, CTO at Interoute, admits that while there’s no tangible outcome from R&D, he’s still going to invest in the best people to do it in-house. “It’s illogical for us to outsource. We have to understand what business we are in,” he says. “If we are building a product, we need to understand how that product goes together. You might struggle to innovate from a cost perspective, but we could never outsource our R&D.”

Geoff Bennett, director of strategic marketing, Infinera, works for a company whose photonic integrated circuit technology relies on cutting-edge research. It supplies products to NTT Communications, Deutsche Telekom and Telefonica, and carries traffic up to 100Gb on a single chip – but it doesn’t use outside suppliers for any core technology. Instead, in spends almost $1 in every $3 of turnover in R&D. That is about double the proportion of Alcatel-Lucent, Nortel or Cisco. It has responded to recent challenges by raising its R&D spend from 20% of turnover in 2007.

“For the core technology, we do it ourselves. In the optical industry, that means the transmission portion of the product. It would be extremely difficult to optimise this with off-the-shelf components,” Bennett says. “The acid test is: does the customer find business value in the feature? If so, it is core to us.” 

Companies can be acquired 

The problem in trusting R&D development to outside companies, he says, is that those companies can be acquired. For an equipment vendor, that’s a real risk, because Cisco is probably the most successful acquirer of innovative companies in the world. The risk was made real to competitors of Infinera who had used Core Optics to do some core R&D: Core Optics was acquired by Cisco in 2010.

“A number of companies decided to license technology from Core Optics, and now the technology they licensed is owned by a competitor. It is no longer a source of competitive advantage to them,” he says.

On the other hand, Infinera is happy to outsource development where there are alternative suppliers of broadly similar products. “There are parts of the product that are not core, for example the software portion is not a key differentiator. So we could have chosen that from a number of companies,” Bennett says. “There are three or four perfectly good sources of that development in the world. That model works well. If our outsourcer disappeared, the product would not stop working overnight, and because we have the source code, we could take over maintenance of that code.”

In this case, the custom development work is done by Metaswitch Networks. “It is really easy to outsource bodies. But this is complex, and for that you need someone who is in the business for the long haul. We assume the people we work with are long-term partners, and we are not just logging some software,” says chairman John Lazar. “We are not a body shop, we develop complex technology. So our customers are not hiring 15 bodies from offshore in order to supplement their R&D teams.” 

The innovation ecosystem 

He sees Infinera’s dividing line used across the industry: just because development work is hard to do doesn’t mean it is core to their business. So carriers and equipment manufacturers alike can outsource underperforming R&D without threatening their competitive advantage.

“Someone like BT is never going to take on an outside company to develop software for a complicated switch. But it may decide to ask someone else to develop the software to glue it into a computer network management strategy,” he explains. “Companies like AT&T have undergone a big shift in the last 10 years, and are now plugging themselves into an innovation ecosystem. For many carriers, it is a response to seeing how the internet has been dominated by companies like Apple and Google, and realising that to respond to this competition, they need to work in the same way.”

At Infinera, Bennett found another way to outsource R&D that may become more common if the explosive growth in blue-sky optical networks research creates companies that are not viable in the long term. “I’m a huge disbeliever in technology acquisition, because everybody who knows anything about the technology tends to leave. In my career it has happened many times,” he says. When Infinera needed to accelerate the R&D of one component of its technology, it acquired the people. “We thought, which is the best company in the world for this? And we decided, Nortel.” At the time, Nortel was in Chapter 11 bankruptcy, and the unit that had done the development was not core to the business long term. Infinera recruited the staff, opening an office near their old workplace so they could retrace their steps. “They can’t copy what they did at Nortel, because that is protected. But it was their first effort. We get their second effort,” he says. “In a way we did outsource the R&D of that product, because somebody else paid for the development of it.”